PASS: A Test Methodology for Cloud Services
While the terminology surrounding cloud services may be a bit cloudy itself, the goal is clear—to create the dynamic data center and infinite scalability, which can reduce the labor-intensive, tactical aspects of providing IT services, freeing managers to pursue innovation and strategic planning. This makes cloud computing testing an imperative.
An IDC survey shows that the top concerns of IT enterprise organizations are performance, security and availability. Cloud services address performance and availability and are offered at different tiers of service:
Infrastructure as a service (IaaS): Hardware (data center space, network equipment and servers), on which the customer installs its own operating systems, applications and data.
Platform as a service (PaaS): Infrastructure with operating system, protocol stacks and development tools installed, on which the customer develops software or hosts servers.
Software as a service (SaaS): Platform with application software installed, on which the customer loads data.
Gartner sums up cloud computing as delivering scalability as a service. Real-time infrastructure (RTI) is a basic building block of the dynamic data center, the fundamental level of IT element automation. It maps the demand for shared services and resources to the supply, making possible the dynamic provisioning and scaling of resources for services to meet service level agreements efficiently and economically. At the higher levels of platform and software services, virtualization can be used to deliver scalability and elasticity by dynamically launching and terminating virtual machine (VM) instances based on varying demand.
Delivering cloud services involves a complex and possibly confounding amalgam of processes, technologies, platforms and devices, many of which are in the early stages of the Hype Cycle. Of course, all systems must be assessed and validated, even those built on mature technologies, and even more so systems that incorporate technology at the early stages of adoption.
Existing technologies have reached maturity and achieved the Plateau of Productivity through the development of test methodologies that assessed and validated specific implementations. The evolving elements of cloud services will do the same only through rigorous testing with new methodologies designed specifically to reveal their weaknesses, and their strengths.
PASS is a test methodology for cloud computing testing that addresses the Performance, Availability, Security and Scalability of cloud services.
Performance: Optimize cloud services and infrastructure to maximize user experience
Ensure high availability in daily operation and under disaster conditions
Eliminate vulnerability and exposure between tenants in the cloud
Scalability: Validate responsiveness as demand varies according to tenant needs
PASS incorporates a wide range of use cases to assess the application components and network elements of a private, public or hybrid cloud. PASS achieves unprecedented granularity through use cases that subject the cloud service to inspection at the infrastructure level (end to end), the inter-server level (applications, services, VMs on different physical servers), and intra-server level (between VMs within a physical server).
For example, at the service level, PASS assesses user quality of experience under realistic and peak subscriber loads, scaling to multi-tenant loads, under the stress of failover scenarios, and while subjected to threats mixed with encrypted traffic.
At the infrastructure level, PASS assesses quality of service performance for the network elements (high-density 10/40/100 Gb Ethernet, converged FC and Ethernet), the application elements (firewalls, intrusion prevention systems, WAN accelerators, proxy servers, and others), and the virtual elements (virtual switches and virtual appliances), end-to-end.
The development of PASS is a milestone in the journey from innovation to adoption, and a significant enabler for those looking to establish world-class cloud services.